Hello dynatrace A.,
In answer to your first question, you have two options. If you would like to edit a specific service to alert then you will have to go to Transactions and Services -> select the service -> click the ellipsis (three dots) -> edit -> select anomaly detection -> uncheck the box to use global settings. Here you can configure your response time degradation automatically or manually. The manual option allows you to alert if your response time exceeds 1000ms (same as 1s). To make this change globally, go to settings -> anomaly detection -> services and make a similar change there.
In regard to your second question, I am not sure. But I would say that the median response time for any request is a good value to go by. The baseline trends for the past 7 days so for any request that you charted you can expand the time frame to 7 days and see where the median values are.
The answer to your third question can be found in the same area as your first. Instead of selecting a manual measure of your response time degradation you will need to select the automatic option. This allows you to track the baseline for your response time and alert based on a certain deviation from it.
Hope that helps,
David Nicholls
